i

NONLINEAR COMPUTATIONAL ANALYSIS OF POSTURAL

STABILITY AND ITS RELATIONSHIP TO ESTROGEN

DEFICIENCY IN POSTMENOPAUSAL WOMEN

by

Zachary Mitchell

A thesis submitted to the faculty of The University of Mississippi in partial fulfillment of

the requirements of the Sally McDonnell Barksdale Honors College.

Oxford

May 2018

Approved by

__________________________

Advisor: Dr. Dwight Waddell

__________________________

Reader: Dr. Matthew Morrison

__________________________

Reader: Dr. Elliott Hutchcraft

ii

© 2018

Zachary Lloyd Mitchell

ALL RIGHTS RESERVED

iii

ABSTRACT

This study aims to determine if multi-scale entropy analysis will more

accurately differentiate the postural stability of postmenopausal women as a

function of their estrogen status than traditional measures of postural assessment.

The study participants included 7 non-Hormone Replacement Therapy (HRT)

postmenopausal women (mean age = 57) and 4 HRT postmenopausal women (mean

age = 55). Postural stability was measured with the NeuroCom Equitest and the use

of conditions 1-4 of the Sensory Organization Test (SOT). The subjects received an

equilibrium score for each condition with a score near 100 representing good

postural stability. A modification of the “Multi-Scale Entropy Analysis” Matlab

program created by Costa et al. was used to analyze the raw center of pressure

(COP) trajectory data measured for each condition in both the medial lateral (x-

direction) and anterior posterior (y-direction) directions. The Matlab program

created complexity index (CI) scores where a high CI is indicative of low postural

stability and a low CI is representative of good postural stability.

A basic t-test of unequal variances was performed on the CI scores between

non-HRT and HRT subjects, in addition to the equilibrium scores between non-HRT

and HRT subjects. The t-test revealed both the NeuroCom Equitest and MSE analysis

identified a significant difference in postural stability between non-HRT subjects

and HRT subjects an equal number of times (n=2), where HRT subjects exhibited

better postural stability than non-HRT. The MSE analysis method was able to

significantly differentiate postural stability between non-HRT and HRT subjects in

SOT 1 (p=0.0432), whereas the NeuroCom Equitest did not. This is of significance

iv

because the subject is standing still on the force plate without having their balance

disrupted. Previous studies have had to manipulate the surroundings and perturb

the subject’s balance to significantly differentiate postural stability between non-

HRT and HRT subjects. The findings from this study indicate that MSE analysis may

provide a higher level of sensitivity necessary to assess postural stability as a

function of estrogen status. The effect of estrogen on the preservation of postural

stability in postmenopausal women may provide further evidence for the

consideration of estrogen as a neuro-protectant.

v

TABLE OF CONTENTS

LIST OF FIGURES…………………………………………….….………vi

LIST OF TABLES………………………………………………….….....vii

INTRODUCTION……………………………………………….…………1

CHAPTER I: TESTING STABILITY………………………………….….5

CHAPTER II: SUBJECTS AND METHODS……………………….…...14

CHAPTER III: RESULTS…………………………………………..….…19

CHAPTER IV: CONCLUSION……………………………………….….21

LIST OF REFERENCES…………………………………………..……..24

vi

LIST OF FIGURES

Figure 1 The Six Conditions of the Sensory Organization Test…….……6

Figure 2 Sensory Organization Test Example Results…………….……..7

Figure 3 The Six Condition of the Motor Control Test…………………..8

Figure 4 Motor Control Test Sample Results…………………………...10

Figure 5 The Relationship Between Age, Variability, and Complexity...11

Figure 6 NeuroCom Balance Equitest…………………………………..15

Figure 7 Center of Pressure Trace………………………………………17

Figure 8 Sample Entropy Scores for Two Different Subjects…………..18

vii

LIST OF TABLES

Table 1 Clinical Score Analysis……………………………………………20

Table 2 Multiscale Entropy Analysis………………………………………20

1

INTRODUCTION

As women reach the ages of 46 to 52 they enter menopause, which is

characterized by a decrease in the sex steroid, estrogen [1]. Recent studies show

postmenopausal women are more likely to fall than their younger counterparts who have

not yet entered menopause [2]. It is considered that the increased risk of falls is related to

estrogen deficiency in postmenopausal women [3]. This estrogen deficiency experienced

by postmenopausal women has been shown to result in low bone mineral density (BMD)

and a reduction in muscle strength [4, 5]. The reduction in muscle strength predisposes

postmenopausal women to falls due to their inability to respond to external stimuli

appropriately [6]. Thus, by combining the effect of a reduction in muscle strength and

low BMD, postmenopausal women become more susceptible to fractures following a fall.

A fall resulting in a fracture places postmenopausal women at a greater risk of developing

further complications that could potentially lead to death [3, 7].

Hormone replacement therapy (HRT) has been introduced as a preventative

measure for falling and fractures by increasing postural stability in postmenopausal

women. Studies have shown that HRT has increased muscle strength and prevented

further BMD loss, which has led to the consideration they contribute to a decrease in falls

[5]. However, despite estrogen deficiency decreasing BMD and subsequently increasing

the risk of fracture upon falling, it has been identified that there is no relationship

between an increase in risk for falls and BMD. Cangussu found there was no statistical

difference between postural stability and the relationship of BMD for postmenopausal

2

women on HRT and BMD for postmenopausal women not on HRT. Therefore, HRT’s

effect on postmenopausal women regarding BMD is a decrease in risk in fracture upon

falling; not a decrease in risk of falling [8].

Despite the positive effects HRT has on BMD and muscle strength, studies have

shown routinely exercising, such as resistance training, has the same results of decreasing

the risk of falling in postmenopausal women [9]. Teixeira et al. concluded HRT showed

no effects on body composition; however, it did protect postmenopausal women from

losing muscle strength. In the study, one group of postmenopausal women received HRT

while the additional group of postmenopausal women did not receive HRT; however,

both groups were enrolled in a one-year resistance-training program. Upon completion

of the program, it was concluded the muscle strength of both groups improved

significantly regardless if HRT was administered [10]. Along with the increase in muscle

strength, women experienced a greater postural balance [9].

Although postmenopausal women have exhibited an increase in postural stability

while performing resistance training, they have shown the same results while only

receiving HRT [11, 12]. Specifically, long-term administration of HRT shortly after the

onset of menopause rapidly increases postural stability to that seen in premenopausal

women [13, 14]. With multiple studies finding HRT maintains postural stability in

women without supplemental exercises; muscle weakness can be ruled out as a key factor

in the risk of falls in postmenopausal women [10-12, 15].

Naessen demonstrated the positive effects of long-term HRT on postural stability

in postmenopausal women when compared to postmenopausal women not using HRT

and concluded the dynamics between HRT and fracture risk were too rapid to influence

3

BMD. Therefore, the increase in postural stability is considered either an effect on the

central nervous system or the preservation of muscle strength, connective tissue and

improvement in mood that lead to an increase in physical activity [11, 12, 16]. However,

since resistance training can preserve muscle strength and connective tissue, estrogen’s

role in maintaining postural stability is best explained as an effect on the central nervous

system [11]. Further consideration of estrogen’s effect on the central nervous system is

supported by its role in the facilitation of the neuronal transmission function [1, 17].

Sufficient levels of estrogen in the blood enable adequate excitation of the sensorimotor

cortex, which is necessary for the maintenance of both the function of motor learning and

coordination. When the level of estrogen decreases to that seen in postmenopausal

women, a decrease in the function of the sensorimotor cortex is observed [17]. This

decrease in function ultimately leads to the insufficient use of the somatosensory and

vestibular systems, thus leading to a decrease in postural stability [11].

Additional studies have examined the relationship between estrogen and postural

stability in the central nervous system; however, they have all only performed linear

analysis on their data to determine the causal effects of estrogen deficiency [11, 17, 18].

To our knowledge, postural stability in postmenopausal women either on HRT or not on

HRT has not been examined with nonlinear analysis. Nonlinear analysis can examine the

variability in recorded data and uncover overlooked information that is not normally

detected using linear analysis [19, 20]. The variability in center of pressure (COP) data

reveals an important condition in the function of motor movements. The function of

variability in human movement used to be considered problematic, however, Stergiou

identified variability as an important characteristic of movement [19].

4

The present study aims to build off the consideration that estrogen’s effect on the

central nervous system plays a vital role in the maintenance of postural stability [11, 12].

Through nonlinear analysis of the variability in movement of postmenopausal women, we

expect to identify a decrease in estrogen in postmenopausal women results in a decrease

in postural stability and an increase in risk of falling compared to postmenopausal women

on HRT. The nonlinear analysis of variability in movement will give insight into the

complexity of the subject’s physiological system. If the subject exhibits a large amount of

complexity, then their physiological system is considered to possess the required

flexibility to adequately respond to external stimuli, thus providing sufficient postural

stability. Whereas, a subject displaying low levels of complexity cannot sufficiently

respond to external stimuli and will, therefore, exhibit low postural stability [21].

According to the “Loss of Complexity Hypothesis,” as individuals age, the complexity of

interactions in their physiological system begins to decrease [22, 23]. Since the average

age of the subjects is n= 56, independent of menopause, they are already experiencing a

decrease in complexity in their system due to the process of aging.

The results of the study are expected to show postmenopausal women on HRT

exhibit a greater amount of complexity in their movement compared to non-HRT users.

The rationale behind this consideration is based on the results that elderly long-term HRT

postmenopausal women reported a better postural balance function than non-HRT users

[11]. Therefore, if HRT users exhibit a better postural balance than non-HRT users, then

their physiological system will possess a larger degree of complexity than that of non-

HRT users [21].

5

CHAPTER I: TESTING STABILITY

Postural stability is measured using the NeuroCom Balance Manager assessment

device through the Sensory Organization Test (SOT) and the Motor Control Test (MCT).

The NeuroCom Balance Manager records the center of pressure (COP) movements of the

subject throughout the tests by utilizing the force platform beneath the subject. The SOT

examines somatosensory, visual and vestibular systems. Whereas, the MCT determines

the subject’s ability to recover from an unexpected disturbance in the force platform.

Together, the results from both the SOT and MCT are interpreted together to depict an

accurate profile of the subject’s postural stability [24].

During the SOT, sway referencing is utilized because the force platform beneath

the patient moves in addition to the surroundings in response to the subject’s

anterior/posterior (AP) sway. Sway referencing refers to the tilting of the force platform

or visual surrounding as the subject’s AP sway fluctuates. Involving sway referencing

requires the subject to rely on their nervous system to more accurately interpret the

sensory inputs from their surroundings. Combining sway referencing with the presence or

absence of eyesight in the subject further increases the subject’s reliance on their

vestibular and somatosensory systems [24].

There are six different conditions in the SOT, however, for this study only

conditions one through four are performed due to their real-world applicability. The four

conditions consist of (1) eyes open and fixed platform; (2) eyes closed and fixed

platform; (3) eyes open and sway-referenced surrounding; (4) eyes open and sway-

6

referenced platform (Fig.1). For each of the four conditions, three trials are performed for

twenty seconds each time. Before each trial, the subject is instructed to stand with their

feet on the designated area on the platform (shoulder width apart), stand quietly, and

maintain their balance. At the start of the trial, the subject is also counted down and given

confirmation that the trial has begun [24].

Figure 1. The six conditions of the Sensory Organization Test [30]

After the completion of each condition, the NeuroCom Balance Manager displays

the results from the three trials in the form of an equilibrium score (Fig. 2). The

equilibrium score is a comparison of the subject’s AP sway during the trial and the

theoretical sway stability limit (12.5 degrees). If the subject’s equilibrium score is near

100, then they experienced minimal sway and adequate postural stability (green bars). A

low score is indicative of the subject swaying to the limit of stability and therefore

experiencing poor postural stability (red bars). The shaded area of the graph is

7

representative of the age-matched normative results. Therefore, a green bar indicates the

subject is either at or above the normative results, and a red bar indicates the subject is

below the normative results [24].

The equilibrium scores for the three trials of each condition are then averaged

together to create an average equilibrium score for conditions 1-4. The average

equilibrium scores are used to identify the subject’s reliance on the somatosensory and

visual systems. The ratio condition 2: condition 1 represents the subject’s reliance on

their somatosensory system to maintain their postural stability. The subject’s reliance on

their visual system to maintain their postural stability is found from the ratio condition 4:

condition 1 [24].

Figure 2. Sensory Organization Test Example Results [30]

The SOT is then followed by the MCT to evaluate the autonomic motor system’s

involvement in maintaining the subject’s postural stability. The MCT consists of six

conditions: (1) small backward translation; (2) medium backward translation; (3) large

8

backward translation; (4) small forward translation; (5) medium forward translation; (6)

large forward translation (Fig. 3). [30] The small translations are indicative of a minimum

threshold, whereas, the large translations elicit an extreme response from the subject.

There are three trials for each condition with a randomized delay of 1.5 to 2.5 seconds

between each trial [25]. The size of each translation is scaled to the subject’s height,

therefore each subject experiences equivalent translational forces in both the forward and

backward direction [24].

Figure 3. The Six Condition of the Motor Control Test [30]

The results from each condition of the MCT are displayed as Weight Symmetry,

Latency, and Amplitude Scaling (Fig. 4). These values are determined independently for

the left leg and the right leg. Similar to the graph of the equilibrium scores for the SOT,

the subject’s scores for Weight Symmetry, Latency, and Amplitude Scaling are compared

to age-matched normative results. The shaded regions in the graphs in Figure 4 indicate

the age-matched normative data to which the subject’s results are compared [24].

9

The Weight Symmetry value represents the subject’s weight distribution in both

the left and right legs. A Weight Symmetry score near 100 indicates the subject had their

weight relatively centered between both of their legs prior to the initiation of the force

platform translation. The Latency score of the subject is the time (in milliseconds)

between the force platform translation and the subject’s active force response. Since

Latency is measured in milliseconds, a low Latency score for the subject is indicative of a

quick active force response in the legs. Analysis of the Latency results reveals a red bar

indicates the subject generated a slower active force response than their respective age

group, whereas, the green bars indicate the subject generated a similar or better active

force response than their age group. The strength of the active force response in each leg

resulting from the translations is represented by the value of Amplitude Scaling. This

value reflects the subject’s ability to produce a response in each leg that is proportional to

the force platform translations. Amplitude Scaling scores representative of sufficient

active force response strength in proportion to the force platform translation will lie

between the shaded regions of the graph (Fig. 4) [24].

10

Figure 4. Motor Control Test Sample Results [30]

For the present study, nonlinear analysis is performed on the results from the SOT

and MCT. The reasoning for using nonlinear analysis on the SOT and MCT results is that

healthy systems exhibit dynamics representative of an extremely adaptable

neuromuscular system. Through nonlinear analysis using entropy techniques, the extreme

adaptability of a healthy physiological system can be identified as complexity in the

system [34]. Lipsitz and Goldberger proposed the “Loss of Complexity Hypothesis,”

which describes that the deterioration of the physiological system is induced by a

reduction in the system’s ability to adequately respond to external stimuli. Specifically,

this hypothesis suggests as physiological functions deteriorate, through aging and disease,

interactions between components in the system break down and cause a reduction in

variability in the system, which ultimately leads to a decrease in complexity in the system

11

(Fig. 5) [22, 23]. It is considered that non-HRT postmenopausal women are experiencing

this decrease in complexity while postmenopausal women on HRT are not exhibiting as

rapid of a decline in complexity as non-HRT [16]. By utilizing analysis of entropy, the

complexity of the physiological system can be used as a representation of the level of

postural stability in postmenopausal women [21].

Figure 5. The Relationship Between Age, Variability, and Complexity [21]

Stergiou performed nonlinear analysis with approximate entropy (ApEn) on COP

data and discovered variability in gait allows the central nervous system to react and

adapt to variable external stimuli one might experience [19]. High values of ApEn signify

data sets recorded at different times do not have similarities, whereas low values of ApEn

signify similarities between the data sets. Therefore, subjects who record a low ApEn

value show a more predictable pattern in their postural stability, which corresponds to

12

both low variability and postural stability. Whereas, higher values of ApEn correspond to

data that do not show similarities, and therefore correspond to both greater variability and

postural stability [19, 26]. Although ApEn measures the predictability of the data, it is

quite sensitive to the length of data [26]. As for SampEn, it has been observed to show

much less sensitivity than ApEn to the length of data it is processing [27]. However, in

both ApEn and SampEn, higher values correspond to greater variability and less

predictability, whereas lower values resemble low variability and greater predictability.

Despite the similarities of ApEn and SampEn in their function, SampEn provides more

consistent results while remaining less sensitive to the data length [26-28].

While the use of ApEn and SampEn provide an insight into the regularity and

predictability of variability, they only consider a one-time scale of measurement.

Analyzing physiological functions at a single time scale does not fully incorporate the

complexity of the entire system. Investigating multiple time scales can detect the

relationship between the various elements of the physiological system that contribute to

the maintenance of postural stability. Therefore, for this study, multiple scale entropy

(MSE) is applied due to its inclusion of multiple time scales. MSE utilizes the SampEn

method by including a coarse-graining procedure, which allows the identification

variations over multiple time scales. The advantages of including multiple time scales are

the complexity can be analyzed at shorter or longer time scales and the measurement of

the overall complexity of the physiological system [21].

Since changes in the physiological system due to aging and disease cannot be

contained to one-time scale, MSE is used to identify the time scales where the

deterioration of the physiological system occurs. In postmenopausal women, this

13

deterioration in the physiological system is attributed to a decrease in estrogen levels in

the blood [16]. Through analysis of MSE on the SOT results, the role of estrogen in the

nervous system’s ability to utilize the sensorimotor and visual systems to maintain

postural stability can be identified. Determining the relationship between estrogen

deficiency and the timescales when postural stability is most affected can provide

valuable evidence for the benefit of HRT in postmenopausal women [21]

14

CHAPTER II: SUBJECTS AND METHODS

The study participants included 7 non-HRT postmenopausal women (mean age =

57) and 4 HRT postmenopausal women (mean age = 55). These women were recruited

through the “UM Today Daily Edition” and flyers passed out to organizations in the

Oxford/Lafayette community. If an individual expressed interest in the study, we

contacted them and conducted a preliminary screening over the phone to verify they were

a candidate for the study. The preliminary screening consisted of (1) Identifying if the

individual is postmenopausal or perimenopausal and (2) If the individual is on or not on

HRT. If the individual identifies as perimenopausal, then they are thanked for their

interest but removed from consideration for the study. An individual identifying as

postmenopausal and on HRT is asked to provide the name of their medication and the

dose. Once the individuals are cleared to participate in the study, they provide written

consent upon arrival to the Neuromechanics laboratory.

The testing apparatus used for the study is the NeuroCom Balance Manager

assessment device (Fig. 6). This device is a commonly used balance testing device found

in assessment clinics. It has a medical insurance number for when a physician orders the

test. The NeuroCom Balance Manager is connected to a computer that actively records

data while each for each test performed.

15

Figure 6. NeuroCom Balance Equitest [30]

The study used a semi-experimental design. The independent variables in the

study were subject height, age, weight and HRT dosage.

When a subject would arrive at the Neuromechanics laboratory, an IRB approved

investigator would provide the subject with a consent form, which they were then

instructed to fill out in its entirety. Upon completion of the consent form, the investigator

would instruct the subject to stand on the force platform of the NeuroCom Balance

Manager. The investigator helped position the feet of the subject in the designated

location on the platform. Once the subject was positioned correctly, and in a comfortable

stance, the investigator proceeded to explain each test the subject will perform (Sensory

Organization Test and Motor Control Test). Before the start of each trial for a condition,

the investigator would count down the subject and begin the test after saying “go.” Upon

completion of each trial, the investigator would instruct the subject to “relax” before the

16

beginning of the next trial. After the completion of both the Sensory Organization Test

and Motor Control Test, the investigator gave the subject a $20 gift card to Wal Mart for

their participation in the study. Upon receiving the gift card, the subject is then instructed

by the investigator to fill out an “incentive payment list” verifying they received a gift

card for completing the study.

Data Collection and Analysis

Although both SOT and MCT data were collected for this study, only SOT data

will be analyzed and presented as the long time series output of this test is amenable for

entropy processing. Complexity Indices for both side-to-side (medial-lateral – ML) and

forward-backward sway (anterior-posterior – AP) COP movement were calculated for

SOT tests 1-4 and compared with clinical output scores one receives from the NeuroCom

unit after each subject is finished with the protocol. Complexity Indices are calculated

from the ML and AP time series which make up the center of pressure traces. An

exemplar center of pressure trace (Fig. 7) is shown for the SOT1 condition.

17

-0.1

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

An

terio

r P

ost

erio

r M

ovem

en

t (M

)

Medial Lateral Movement (M)

Center of Pressure Trajectory

Figure 7. Center of Pressure Trace

A Matlab program was created which was a modification of the “Multi-Scale

Entropy Analysis” program, originally created by Costa et al., (2015) through the NIH

administered PhysioNet website [29]. The maximum timescale Tau was set to 10 which

establishes the granularity of the analysis. The program “creates a set of coarse-grained

series with a user-selected range of scales and calculates SampEn for each coarse-grained

series.” A complexity Index (CI) is calculated by adding all of the SampEn scores across

all Tau values (area under the curve) as shown in exemplar Fig. 8 [30].

We hypothesize that MSE scores will more accurately differentiate the postural

stability of post-menopausal women as a function of their hormone status. Basic t-tests

(unequal variances) were used to test this hypothesis.

18

Figure 8. Sample Entropy Scores for Two Different Subjects

19

CHAPTER III: RESULTS

The results from the NeuroCom Equitest show HRT subjects exhibit a larger

mean equilibrium score than non-HRT subjects, which is indicative of greater postural

stability (Table 1). The MSE method created CI scores in both the medial-lateral (x) and

anterior/posterior (y) directions where a low CI is representative of good postural

stability. In Table 2, it is evident that HRT subjects display a lower CI than non-HRT

subjects in all conditions except for SOT 1y and SOT 4y. Therefore, both methods of

analysis were able to produce results consistent with those found in previous studies [11,

12].

Further analysis of the traditional clinical scores and complexity indices reveals

both methods determined there to be a significant difference in postural stability between

postmenopausal women on HRT and non-HRT postmenopausal women an equal number

of times (n=2). The NeuroCom assessment device identified a significant difference

between HRT and non-HRT postmenopausal women in SOT 2 (p≤ 0.05) and SOT 3

(p≤ 0.05) (Table 1). Whereas, the MSE method determined there to be a significant

difference between HRT and non-HRT postmenopausal women in SOT 1x (p≤ 0.05)

and SOT 4x (p≤ 0.05) (Table 2).

20

Table 1. Clinical Score Analysis

Test Mean Equilibrium

Score for Non-HRT

Mean Equilibrium

Score for HRT

P Score Significance

SOT 1 92.543 95.083 .0974 No

SOT 2 89.381 93 .0050 Yes

SOT 3 86.619 91.667 .0330 Yes

SOT 4 77.619 85.75 .0674 No

Table 2. Multiscale Entropy Analysis

Test Mean Complexity

Index for Non-

HRT

Mean Complexity

Index for HRT

P Score Significance

SOT 1x 13.802 13.064 .0432 Yes

SOT 1y 14.605 14.838 .4197 No

SOT 2x 14.245 13.530 .1153 No

SOT 2y 15.969 15.537 .3170 No

SOT 3x 13.864 13.173 .0562 No

SOT 3y 15.527 15.157 .3306 No

SOT 4x 14.369 13.657 .0236 Yes

SOT 4y 15.723 16.074 .2823 No

21

CHAPTER IV: CONCLUSION

The results from both the NeuroCom and MSE method were consistent with the

findings from previous studies in showing that postmenopausal women on HRT exhibit

better postural stability than non-HRT postmenopausal women [11, 12]. The MSE

analysis for SOT 1y and SOT 4y showed non-HRT subjects have greater postural

stability than HRT subjects, however, these results may be due to the limited availability

of subjects on HRT in the present study. Despite SOT 1y and SOT 4y, graphical analysis

of all the SampEn scores across all Tau values for the remaining SOT conditions

indicates HRT subjects exhibit lower SampEn scores and thus a high degree of

complexity in their COP movement (Fig. 8). These findings support the “Loss of

Complexity Hypothesis” in showing that HRT subjects exhibit less of a breakdown of

complexity in their movements compared to non-HRT subjects [17, 23].

The present study sought to use MSE as a method to more accurately assess the

postural stability of postmenopausal women with regards to the presence/absence of

estrogen in their system. A t-test performed on both the NeuroCom results and MSE

results indicated that both methods accurately detect a significant difference in postural

stability between HRT and non-HRT postmenopausal women an equal number of times

(n=2). NeuroCom analysis found a significant difference in SOT 2 and SOT 3, however,

since the subject is blinded in SOT 2 and their surrounding is sway-referenced in SOT 3,

we expected the NeuroCom to detect a difference in the subjects. Previous studies have

had success in differentiating postural stability in non-HRT and HRT subjects, however,

22

they have only done so by creating conditions for the subjects similar to those found in

SOT 2 and SOT 3 [12]. For the present study, we are interested in determining if the

MSE method can detect a difference in the subjects without needing to manipulate the

conditions to elicit a fall.

When the MSE method was applied, a significant difference was identified in

SOT 1x and SOT 4x. Although the MSE method was unable to detect more differences

than the NeuroCom, it was able to identify a significant difference in SOT 1, whereas, the

NeuroCom was not. To our knowledge, this is the first instance to identify a significant

difference in the postural stability of non-HRT and HRT postmenopausal women

experiencing the conditions present in SOT 1. Previous studies have not shown these

results because their use of traditional assessment methods does not possess the required

sensitivity to differentiate an HRT subject and a non-HRT subject without trying to elicit

a fall in the subject [12]. Therefore, applying the MSE method to analyze COP data may

introduce a higher level of sensitivity necessary to assess postural stability as a function

of estrogen status.

MSE may be more likely to detect the complex interactions between the sensory,

motor and central processing systems involved in postural stability, however, prospective

studies are needed to show if the proposed method is more accurate than traditional

methods. Currently, if used in conjunction with the NeuroCom, the MSE method can

create a more accurate representation of a postmenopausal woman’s postural stability as a

function of the presence of estrogen than previous traditional methods. Thus, we can

apply the results from both methods of analysis to support the assumption presented by

Naessen et al that estrogen can counteract the deterioration of complexity in movement

23

through an effect on the central nervous system. Additional research on the MSE method

and its ability to analyze postural stability in postmenopausal women may provide further

evidence of estrogen’s role as a neuroprotectant and protective factor for risk of falling.

24

LIST OF REFERENCES

1. Sipilä, S., T. Finni, and V. Kovanen, Estrogen Influences on Neuromuscular Function in Postmenopausal Women. Calcified Tissue International, 2015. 96(3): p. 222-233.

2. Haber, N.E.L., et al., Relationship between age and measures of balance, strength and gait: linear and non-linear analyses. Clin Sci, 2008. 114.

3. Rosner, C., Menopause Puts Women At Higher Risk Of Fall-Related Injuries. 2017, WNPR: wnpr.org/post/menopause-puts-women-higher-risk-fall-related-injuries.

4. Arnold, C.M., et al., The relationship of intrinsic fall risk factors to a recent history of falling in older women with osteoporosis. J Orthop Sports Phys Ther, 2005. 35(7): p. 452-60.

5. Heikkinen, J., et al., HRT and exercise: effects on bone density, muscle strength and lipid metabolism. A placebo controlled 2-year prospective trial on two estrogen-progestin regimens in healthy postmenopausal women. Maturitas. 26(2): p. 139-149.

6. Phillips, S.K., et al., Muscle weakness in women occurs at an earlier age than in men, but strength is preserved by hormone replacement therapy. Clinical Science, 1993. 84(1): p. 95-98.

7. Prieto-Alhambra, D., et al., An increased rate of falling leads to a rise in fracture risk in postmenopausal women with self-reported osteoarthritis: a prospective multinational cohort study (GLOW). Annals of the Rheumatic Diseases, 2012.

8. Cangussu, L.M., et al., Evaluation of postural balance in postmenopausal women and its relationship with bone mineral density- a cross sectional study. BMC Musculoskeletal Disorders, 2012. 13(1): p. 2.

9. Charette, S.L., et al., Muscle hypertrophy response to resistance training in older women. Journal of Applied Physiology, 1991. 70(5): p. 1912-1916.

10. Teixeira, P., et al., Resistance training in postmenopausal women with and without hormone therapy. Medicine and science in sports and exercise, 2003. 35(4): p. 555.

11. Naessen, T., B. Lindmark, and H.-C. Larsen, Better postural balance in elderly women receiving estrogens. American Journal of Obstetrics & Gynecology. 177(2): p. 412-416.

12. Naessen, T., B. Lindmark, and H.C. Larssen, Early postmenopausal hormone therapy improves postural balance. Menopause, 2007. 14.

13. Ekblad, S., et al., Disturbances in postural balance are common in postmenopausal women with vasomotor symptoms. Climacteric, 2000. 3(3): p. 192-198.

14. Figueroa, A., et al., Effects of exercise training and hormone replacement therapy on lean and fat mass in postmenopausal women. J Gerontol A Biol Sci Med Sci, 2003. 58(3): p. 266-70.

25

15. Ribom, E.L., et al., Six months of hormone replacement therapy does not influence muscle strength in postmenopausal women. Maturitas. 42(3): p. 225-231.

16. L. Hammar, M., et al., Effects of hormonal replacement therapy on the postural balance among postmenopausal women. Obstetrics & Gynecology, 1996. 88(6): p. 955-960.

17. Smith, S.S., Estrogen administration increases neuronal responses to excitatory amino acids as a long-term effect. Brain Research, 1989. 503(2): p. 354-357.

18. Bergstrom, I., B.-M. Landgren, and I. Pyykko, Training or EPT in perimenopause on balance and flushes. Acta Obstetricia et Gynecologica Scandinavica, 2007. 86(4): p. 467-472.

19. Harbourne, R.T. and N. Stergiou, Movement Variability and the Use of Nonlinear Tools: Principles to Guide Physical Therapist Practice. Physical Therapy, 2009. 89(3): p. 267-282.

20. Peng, C.K., Fractal mechanisms in neuronal control : human heartbeat and gait dynamics in healthy and disease. Self-Organized Biological Dynamics & Nonlinear Control, 2000: p. 66-96.

21. Busa, M.A. and R.E.A. van Emmerik, Multiscale entropy: A tool for understanding the complexity of postural control. Journal of Sport and Health Science, 2016. 5(1): p. 44-51.

22. Lipsitz, L.A., Dynamics of StabilityThe Physiologic Basis of Functional Health and Frailty. The Journals of Gerontology: Series A, 2002. 57(3): p. B115-B125.

23. Lipsitz, L.A. and A.L. Goldberger, Loss of complexity and aging: Potential applications of fractals and chaos theory to senescence. JAMA, 1992. 267(13): p. 1806-1809.

24. Neurocom. NeuroCom. Objective Quantification of Balance and Mobility. . 25. Vanicek, N., et al., Computerized Dynamic Posturography for Postural Control

Assessment in Patients with Intermittent Claudication. Journal of Visualized Experiments : JoVE, 2013(82): p. 51077.

26. Richman, J.S. and J.R. Moorman, Physiological time-series analysis using approximate entropy and sample entropy. American Journal of Physiology-Heart and Circulatory Physiology, 2000. 278(6): p. H2039-H2049.

27. Ramdani, S., et al., On the use of sample entropy to analyze human postural sway data. Medical Engineering and Physics. 31(8): p. 1023-1031.

28. Smith, B., et al., Sample Entropy Identifies Differences in Spontaneous Leg Movement Behavior between Infants with Typical Development and Infants at Risk of Developmental Delay. Technologies, 2017. 5(3): p. 55.

29. Costa, M., A.L. Goldberger, and C.K. Peng, Multiscale entropy analysis of biological signals. Physical Review E, 2005. 71(2): p. 021906.

30. Costa, M., A.L. Goldberger, and C.K. Peng, Multiscale Entropy Analysis of Complex Physiologic Time Series. Physical Review Letters, 2002. 89(6): p. 068102.